DIY Heat Bank: PHE specifications

[Mike Holmes]

I am shortly going to convert my conventional open vented DHW cylinder into a Heat Bank following the design given at

am going to set the post-PHE TMV to distribute DHW at 55degC. What I need to know is how I go about determining the required PHE secondary output temperature. Say the PHE output is designed to be

60degC. If one shower is running, for example, the TMV will blend a little cold water to achieve 55degC. That's easy. What I'm finding difficult is how to be sure what happens when the 2nd shower is switched on, and then perhaps the kitchen hot tap...

I have a very good supply from the rising main. It runs at ~8bar (reduced to 5bar by a pressure reducing valve at entry into the house) and delivers 30l/min at 8bar.

When additional outlets ae opened, will the secondary output temperature will drop as the flow rate through the PHE increases? Or perhaps it won't... perhaps the flow rate to each outlet will drop... and the temp will stay the same...?

Basically what I don't quite understand is the dynamics between water pressure, flow rate and number of outlets open.

Many thanks in advance for any info on this.

Mike

[Mike Holmes]

mmm... probably goes without saying but in my original post, I didn't add that I want to ensure the delivered water is not less than 55degC, even if 2 showers and a hot tap are open simultaneously. So that means I need to be able to come up with a required PHE secondary output temp for my design, so I can spec the heat load/performance characteristic for the PHE.

Cheers

Mike

[John Rumm]

One way would be to work out the total energy requirements at the point of delivery. Lets say you want the final mix temperature at 42 degrees and your worst case ground water temp is 5 degrees...

Vc[42 - 5] = Vh[55 - 42]

so your mix ratio is 37:13 hot to cold. If you main can supply 30 lpm, and you could use that full capacity, to supply water at 42 would mean you need 22.2l of hot per min, and 7.8l of cold

So you want 22.2 lpm at 55 deg, and the worst case ground water temperature is 5 degrees. That gives a 50 degree delta. So 4200 x 50 x

25 = 4.66MJ / min or 77.7KJ/sec or 77.7kW.

So a 100kW PHE sounds like it should cope with ease.

Note that if you have a 300L store at 75 degrees, and the PHE minimum differential is say 5 deg that only give you 15 degrees, to play with or

18.9MJ or about 4 mins. With real time replenishment from the boiler you may extend that to 6 min. So long as the showers are thermostatically controlled however you an cope with a much greater fall in store temperature. Obviously the more modest the shower flow rate requirements the better, and the more boiler power available the better.

[Mike Holmes]

Very good point about the capacity requirement for a high flow rate like this - thanks for the heads-up on that. I shall have to invest in a higher capacity tank for sure. My existing one is just

96litres! Good opportunity to get a direct cyclinder suited to the job anyway... and the total cost is still a tiny fraction of the Pandora equivalent... and only a little more than a decent pump which can deliver a similar performance hot water pressure and flow rate... which happens to use about 1.5kW of electricity when operational!!

========

**** Also - can anyone help on the below (see the original post for the full explanation): ****

Many thanks

Mike

[Richard Skeen]

possibility of eliminating the thermostatic mixing valve.

As far as I can see, the DPS Pandora heatbank has a flow switch to control the pump and a TMV to regulate the output temporature, whereas the Gledhill Pulsacoil heatbank doesn't have a TMV. Instead, it has a temperature sensor on the DHW side and a variable speed pump.

What do people think about using a variable speed pump and temperature sensor instead of a TMV? Would this approach would also give better stratification inside the thermal store? How easy is it to find a variable/multi-speed circulator pump?

Rich.

[Mike Holmes]

You'd need a flow switch capable of detecting different flow rates hooked up to some sort of controller to deliver variable current to your pump. I guess you could then use a standard pump.

Mike

[Mike Holmes]

Meant to add - if you reckon you'll draw a lot of your DHW off at lowish flow rates then this should preserve stratification much better, but then when you use the shower - presumably at full bore, or anything else eg hot ll washing machines, then it won't make any difference

HTH

MIke

[Mike Holmes]

Meant to add - if you reckon you'll draw a lot of your DHW off at lowish flow rates then this should preserve stratification much better, but then when you use the shower - presumably at full bore - or anything else eg hot fill washing machines, then it won't make any difference

HTH

MIke

[John Rumm]

More open outlets increases flow rate, and reduces dynamic pressure.

[John Stumbles]

Good idea. Yes. Not easy at all :-(

[dom]

I think most pumps use squirrel cage rotors, so a variable frequency drive (vfd) could be used to give variable speed on a standard pump:

aren't cheap, but the upside is they have lots of control inputs

- so sensor hook-up would probably be straightforward.

A 150W VFD would probably be sufficient (this isn't the right one, but it gives you the idea):

'd need both a flow switch and a temperature sensor. Probably just hook the flow switch to an interlock on the vfd - and an appropriate temperature sensor (that has a linear voltage output) to the external speed adjustment on the vfd (which can usually accept an external pot or something like an external 0-10Vlevel).

Finally you would need to achieve the appropriate control action - with the right choice of VFD it may be possible to do that in the firmware. In my experience of using a VFD to run a 3 phase bandsaw, I found my Omron unit had everything I could possibly need and lots more besides. They're designed to be ultra-flexible for a huge spectrum of uses.

Perhaps someone else would like to comment on setting an appropriate control action - vfd's usually offer PID control:

'll probably find there's also lots of manufacturer info about hooking up temp sensors to vfd's for industrial processes.

[Doctor Drivel]

Read

for an explanation of how heat banks work

1. Fit a Surrey type of flange in the top tapping of cylinder. 1" if you can.
2. The plate heat exchanger connects to the Surrey flange port that is immersed in the cylinder water. Drill a few holes in the protruding tube that goes into the water to ensure heat is extracted from cylinders dome, not being pumped to the bottom of the cylinder. The protruding tube has a plate over the end to stop the water heat right to the bottom of the cylinder.
3. The boiler flow is taken from the other connection of the Surrey flange on the top of the cylinder. The return into the old cold feed tapping.
4. From the Surrey to the plate and then to the pump the to a spring loaded check valve and back into the bottom cylinder port. Make a diffuser by inserting 22mm pipe into the bottom cylinder port and drilling holes in and stopping up the end with an internal pipe stop (available from BES). File down the inside of a brass compression fitting that screws to the cylinder port, removing the pipe stop so the pipe can go straight through. This will spread the returning water mainly down, so it will not upset stratification in the cylinder.
5. Two cylinder stats can be used to give a long efficient boiler recovery burn. One stat about half way up and the other about 25% of the way up the cylinder. Set bottom say to 70-80C, set top say to 60-65C. The stats must be latched in with a relay.
6. The cold mains direct from the cold mains stoptap with no tee offs. Take into the flow switch then into the bottom connection of the plate heat exchanger and then to the DHW blending valve.
7. Have thermostatic shower mixers and take the hot supply for these directly off the plate heat exchanger "before the TMV (blending valve). No need to run DHW through two thermostatic mixers.
8. Install a phosphor de-scalar on the incoming mains pipe.
9. Install isolation valves on heavy usage appliances such as the washing machine, and throttle back so it will not rob showers and baths of hot water.
10. Have the F&E tank top up at the bottom of the cylinder and vent at the top. You may want to vent from the boiler flow pipe.
11. Fill with inhibitor - about 1% of total system volume. An average system is approx 100 litres, so a system has one 1 litre can. If say 150 litre cylinder then two cans for the cylinder alone, which is three.
12. Fit a Magnaclean filter on the rads return pipe to the boiler.

The performance is brilliant and you will be delighted with the mains fed showers and no vibrating power shower pump noise. High pressure mixers can be on all appliances.

Gledhill will supply a Plate Heat X. The model for the 145 litre Systemate will do. If you can get another cheaper source then try them. A 100kW plate heat exchanger is needed

or Screwfix will supply most of the fittings. They don't do the plate heat exchangers.

Farnell will supply the flow switch

Farnell number:

1006771 22mm compression joints.

Flow Switch, makers site: The FS06

flow switch is about the best - very good.

If the boiler requires to be in a sealed system then have a cylinder with a quick recovery coil an the boiler heats this ASAP. Most boilers can be fitted to an open vented system.

[Doctor Drivel]

Best see this thread:

a mixing valve on the primary water sensing DHW temp on the secondary water. This will only alllow water into the bottom of the cylinder at the rate which is required - the minimum rate preserving stratification. Speader pipes will assist too.

[Doctor Drivel]

Differential at 5C? Are you serious?

Have two tappings at the bottom and run the CH off this. Have TRVs on all rads and a use a Grundfoss Alpha on the CH circuit. Then no room t/stat needed.